Emerging Trends in Drugs, Addictions, and Health (Dec 2024)

Linking online activity to offline behavior: A meta-review of three decades of online-to-offline scholarship with future implications for AI

  • Scott Leo Renshaw,
  • Kathleen M. Carley

Journal volume & issue
Vol. 4
p. 100154

Abstract

Read online

As society grapples with the emerging significance and implications of Large Language Models (LLMs), such as OpenAI’s ChatGPT, or Google’s Gemini, as well as other advancements in modern generative Artificial Intelligence (AI), it is crucial to recognize the existing role that data, algorithms, and online social networks have already played in shaping our contemporary society. This review article provides the first comprehensive examination of the current state of knowledge, across disciplinary divides, on how online influences impact offline behaviors, laying the necessary groundwork for investigating and researching the potential impact that these new technologies will have on our “offline” lives. Through a deep-dive collection of articles (n=149), we review and analyze research with measurable Online-to-Offline impacts (n=88). Within this Online-to-Offline criteria, we identify five emergent cross-cutting themes, namely: Social Diffusion, Social Reinforcement, Social Boundary & Identity Maintenance, Cognitive and Attitudinal Research, and Research on Vulnerable & Marginalized Impacts. Through a second wave snowball collection process, we construct a citation network from the broader Online and Offline research literature, allowing us to locate the Online-to-Offline subset as part of a larger intellectual discussion. Finally, we conduct a Term Frequency-Inverse Document Frequency (TF-IDF) analysis of terms used in the titles of these online/offline research papers, from 1990 to 2023, to identify the evolution of researchers’ conceptualization and framing of Online and Offline research across the past 30 years. The meta-review, presentation of high-level cross-cutting interdisciplinary themes, co-citation network analysis, and TF-IDF analysis collectively provide a cohesive and deeper understanding of the research space of online/offline influences. By taking stock of the ways in which online factors have already shaped individual, group, or organizational behaviors and social dynamics broadly in “offline” contexts, this work aims to provide a cohesive theoretical and empirical foundation for future researchers to better anticipate, address, and frame the future consequences of the rapidly evolving digitally influenced landscape we find ourselves in today.

Keywords